inya.bilingualmanualsocialresearch methods.pdf
TRANSCRIPT
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
1/359
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
2/359
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
3/359
Authors:
Maxime Boutry, a social anthropologist (Ph.D. EHESS - Paris) specializing on Myanmar, has been Lead
Researcher for a number of development projects since 2008. His academic publications include:From
Padi States to Commercial States (Amsterdam University Press, 2015) as co-editor.
Tun Myint is Assistant Professor, Faculty of Health Sciences, Simon Fraser University, Barnaby, British
Columbia, Canada. He is specialized in Global Health, Program Planning and Evaluation, and
Quantitative Research Methods.
Franois Tainturier, Ph.D. (SOAS, London) is the Inya Institutes Executive Director. He has eight
years of experience in developing and coordinating research and training programs in the field of social
sciences and humanities first in Cambodia (2001-2005) and since 2011 in Myanmar.
Translators: Naing Tun Lin - Ye Wunna - Wai Phyo Maung
English version copy editor: Myanmar version copy editor:
Charles Carstens San San Hnin Tun, Ph.D.
Inya Institute (www.inyainstitute.org) is a Yangon-based higher-learning institute dedicated to advancing
knowledge in the social sciences, arts and humanities in the field of Myanmar Studies. The institute is a
non-political, non-religious, non-profit, and non-degree-granting organization. It has the following goals:
(1) to build research capacity for young local researchers; (2) to foster scholarly exchange between local
and international researchers; (3) to encourage a critical approach to scholarship; (4) to raise the publics
understanding of the countrys multi-faceted cultural heritage.
Project and Publication generously funded by:
and supported in part by a grant from the Open Society Foundations
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
4/359
(Ph.D. EHESS -)
From Padi States to Commercial
States()
Ph.D. (SOAS) Inya Institute
(-)
PhD
Inya Institute (www.inyainstitute.org )
() () () ()
Open Society Foundations
http://www.inyainstitute.org/http://www.inyainstitute.org/http://www.inyainstitute.org/http://www.inyainstitute.org/ -
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
5/359
Table of Contents
Foreword 1Introduction 7
Glossary 19
Basic 1. Defining a Research Problem 27
Basic 2. Quantitative and Qualitative Research 39
Basic 3. Baseline Study 58
Basic 4. Survey Research (1) 74
Basic 5. Survey Research (2) 84
Basic 6. Sampling: Non-Probability Sampling 98
Basic 7. Sampling: Probability sampling 110
Basic 8. Focus Group Discussion (FGD) 124
Basic 9. In-depth Interviews 144
Basic 10. Participatory Learning Appraisal and its Tools 160
Basic 11. Participatory Action Research (PAR) 184
Basic 12. Research and Ethics 204
Intermediate 13. Research and Project Design 222
Intermediate 14. Case study 240
Intermediate 15. Quantitative Data Analysis 254
Intermediate 16. Qualitative data analysis 276
Intermediate 17.Using Secondary Sources or Existing Data 292Intermediate 18.Evaluation Research 300
Intermediate 19.Supporting Details 314
Intermediate 20.Writing a Report 330
Intermediate 21. Policy and Advocacy brief Writing 342
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
6/359
- - - () - () - -() - - - - - - - - - - - -
- - - - -
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
7/359
1
Foreword
There is a shared agreement among local and international organizations as well as
experts in the field that social projects developed in Myanmar lack the following three
components:
(1) Research that helps to gain better knowledge of the communities targeted by
social projects;
(2) Monitoring that helps to assess progress made in the implementation of a
project based on its initial plan;
(3) Evaluation that helps to assess how a completed project has reached its
objectives.
Responding to this shared agreement, the present manual introduces Myanmar
readers to social research methods that can improve the implementation of socialprojects in the country, particularly regarding the three components mentioned
above.
The manual is primarily targeted to Myanmar individuals who are involved in social
projects or have an interest in developing their research skills, but have little
knowledge of how to conduct social research. We also hope that this manual will be
of interested to a larger audience at a time when social research is much needed in
Myanmar.
The manual is divided in two parts: Basic level (Chapter 1-12) for those who have noprior knowledge of how to conduct research; and Intermediate level (Chapter 13-21)
for those who have some acquaintance with social research. Chapters in the first part
introduce the basic concepts of social research and the research methods commonly
used for collecting data. Chapters in the second part elaborate some of the concepts
introduced in the first part. They also give guidance on how to produce evidence
based on the data collected and how to produce reports.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
8/359
2
-
()
()
()
()
( )
( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
9/359
3
In terms of content, emphasis was placed on writing this manual in an accessible
and concise way. We avoided technical jargon as much as possible for two reasons:
first, vocabulary related to research methods in both the English and Myanmar
languages is only understood by a very small minority of people in Myanmar;
second, there is not yet a widely-accepted agreement on technical terms and theirequivalent in both languages. Whenever needed, a possible Myanmar-language
equivalent was suggested. Some of these suggestions, however, may not be widely
accepted. Time and more practice of social research in Myanmar will eventually
bring greater currency to some equivalents rather than others.
Along these same lines, we aimed to include concrete examples that are related to
the Myanmar context. The value of examples as learning tools has long been
recognized. Learners can more easily grasp explanations about concepts or ideas if
they can associate them with experiences or contexts with which they are familiar.
Lastly, one of the main features of the manual is the English- and Myanmar-
language texts placed face-to-face. We believe that this page layout is in itself a
very effective pedagogical tool. It is often the case that Myanmar readers read
English-language texts first and believe they have fully grasped their meaning. This,
however, may not be necessarily the case. With the two versions placed side by
side, readers will be able to confirm whether or not they have correctly understood
the text.
The manual is an introduction to social research methods and focuses mainly on
applied research. This means that it offers practical tools on how to conduct
research, report findings, and, ultimately, give recommendations so as to improve
the lives of those in the communities under study. Because of this emphasis, we
deliberately omit many of the theoretical issues that are usually discussed in
textbooks about social research methods. Such issues may not have direct
relevance for those who have little research experience and expect practical
guidance. But it will of great benefit to discuss these issues when the teaching of
social research methods reaches a more advanced level in Myanmar.
With the country opening up, it is often said that international assistance ought to
disseminate the benefits of critical thinking as part of the current efforts to
reinvigorate Myanmars education system. It is indeed a very important issue. Yet
this is one that may not be an immediate priority. What seems more pressing is to
provide learners with tools that help structure their thinking. What seems more
critical at this stage is to guide these learners to use these tools so that they may
link observations and data collected from the field research to broader reasonings.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
10/359
4
-
-
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
11/359
5
Further implementation of international programs in Myanmar requires
observing commonly accepted international procedures and standards.
We hope that the present guide will help readers understand what these
procedures and standards are and how they may be followed.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
12/359
6
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
13/359
7
Introduction: What is social science research? What is its purpose? How is it done?
General definition: Social science research (hereafter referred to as social
research) means studying in a scientific way how people act and interact in a
society. In order to study these issues, researchers usually study how people
think, behave, and act in a particular context. People here means either
people as individuals or groups of individuals, or people as embodied in
institutions, associations, or any types of formal or informal organizations.
The social study of people is an activity that aims to produce a correct, true,
and scientific understanding of a situation or problem involving individuals or
groups of people. Researchers attempt to understand this situation or
problem through direct experience.
Producing this correct, true, and scientific understanding requires working
systematically and rigorously based upon observations and data collected
during the research. Observations and data are collected by researchers who
use social research methods that have been long recognized by social
researchers. The researcher then interprets both observations and data and
uses them as supporting details or research findings. The end result is a report
that presents the researchers understanding and analysis of a particular
issue.
Purpose of social research: Social research aims to contribute to existing knowledge
and to generate new knowledge. It is important that this knowledge is shared and
disseminated to a broad audience. In doing so, social researchers participate in the
general progress of the society. In the case of social research applied to communities
in need, researchers also formulate recommendations which address social problems
affecting these communities.
Types of social research: For clarifying purposes, social research is often divided into
the following three types of investigation:
An exploratory type: this type of investigation is only concerned with topics
that have largely remained underexplored.
A descriptivetype: this type of investigation is concerned with a simple
account of observations and data.
An explanatorytype: this type of investigation is about establishing a
relationship between two situations: (1) a relational link and (2) a causal link.
The relational link shows a relationship between two situations and, possibly,
how the significance of this relationship. The causal link shows how specificfactors from one situation influence the other, or generate a problem.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
14/359
8
( )
[]
()
()
()
()
-
()()
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
15/359
9
The boundaries between these three main types of investigations, however,
are not strict. It is often the case that a research project may blend at least two
of these three types of investigations. Explanations on these three types will
receive further development in the first chapter.
Among all types of investigation above mentioned, it is probably the
explanatory type which is the most challenging.
One of the main problems of the explanatory type is confusion between
relational and causal links. In English-language textbooks, this confusion is
summed up as follows: correlation does not implycausation. Simply because
one event follows another or two factors vary at the same time does not mean
that one causes the other (causal link) or that they are even linked (relational
link). It is one of the benefits of research to identify a possible link between
two events, factors, or situations and to clarify possible confusion.
Explanatory investigations also lead to a frequent misconception about causal
links. This misconception is the following: because a causal link is identified
between two events, two factors, or two situations does not mean that
researchers have a strong basis upon which to elaborate a broader pattern
between events, factors, or situations of the same nature. When trying to
explain a causal link, it is sometimes difficult to access sufficient information toprecisely understand why a particular situation or problem arose. This
limitation makes it often impossible to observe and identify all causes within
the time of the research process. This means that researchers have sometimes
to infer causes. In other words, they have to draw a conclusion about why
something happened by thinking logically and rationally and based on
assumed facts. In conducting social research, it is critical to ensure that
inferences that are not valid are avoided. This is achieved when (1) the process
by which researchers infer causes is logical and rational, and (2) the facts that
are assumed by the researchers are likely to have occurred. All this is to show
that researchers have to remain cautious about establishing causal links. In
other words, it is important to be alert to the danger of elaborating a broader
social pattern on the basis of a causal link.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
16/359
10
[]
()
() () ()
()
()
() ()
[]
()
()
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
17/359
11
Another important issue for explanatory investigation is the pitfall
of simplistic explanations about causal links. Simplistic explanations
about why something happened are those which give more weight
to determinism than probability. Causal links are explained in the
same way as the rule of nature governs things and beings in the
field of natural sciences: for instance, whenever water is heated to
100C, it always boils. This causal link is acceptable for the field of
natural sciences. But, in the field of social sciences which studies
human action and interaction, this type of explanation cant be
accepted. Causal links do not invariably lead to the same outcome
because human behavior is complex. This is true even when pre-
existing conditions appear the same. Therefore causality in the
social sciences is more about probability than about determinism.
In other words, researchers can say a factor or a set of factors
increases (or decreases) the probability of a particular outcome.But researchers cant say that this factor or set of factors will
inevitably lead to this particular outcome.
This overall approach to causalityboth more probabilistic and
sensitive to contextwill hopefully leads to a better understanding
of a particular problem or issue. It may also raise researchers
awareness that a problem should not be thought in black or
white terms only. That is, a problem is better approached by
trying to identify the varying shades of grey of social behavior.
When researchers study social behavior, they have to develop theirtolerance for ambiguity. This means to take into account
someones ability to have multiple and possibly conflicting views in
his/her mind. These multiple and possibly conflicting views are
simultaneously upheld by the person who does not deny any of
them. One of the tasks of the researcher is to report about all of
them without making any judgment.
Scientific process and social research: a scientific understanding of
a problem or issue has two main characteristics: (1) it has to make
sense and follows some logical thought; and (2) it is consistent withall the research observations.
In everyday life, people sometimes reason illogically. By contrast, as
researchers structure their thinking, they have to come up with a
logical reasoning. Also in everyday life, people often observe
inaccurately and make very general conclusions based on only few
observations. By contrast, researchers make careful observations.
They also make large number of observations. This large number
ensures that the conclusions reached by researchers take into
account what was learned from all observations.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
18/359
12
C
() () ()
()
()
()
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
19/359
13
Developing a scientific understanding of an issue based on these two characteristics can
be done in two ways: an inductive approach and a deductive approach.
The inductive approach means that the research process moves from the particular to the
general. The researcher uses the observations collected during the field research. The
researcher then tries to identify a social pattern based on a number of observations as
large as possible. To identify a social pattern means to identify some regularity and order
in all the observations made. This social pattern is linked to a larger reasoning about why
the issue or problem emerged. When linking this pattern to a larger reasoning,
researchers produce some generalizing principles about the issue or problem. This
inductive approach is the way that researchers with little experience frequently use.
The deductive approach means that the research process moves from the general to the
particular. The researcher tests concepts and patterns known from theory. Before testing
these concepts and patterns, the researcher formulates a hypothesis that will be
confirmed or rejected by research findings. A hypothesis is a statement of what the
research expects to observe in reality when conducting the research if the theory is
correct. After collecting observations and data, the researcher determines whether the
results are consistent with the initial hypothesis. If research findings do not support the
initial hypothesis, this should not be seen as a problem at all. It simply shows that research
was critical in clarifying some misconceptions made about the issue or problem
addressed.
These two approaches, inductive and deductive, are the two most frequent approaches of
conducting research. It is usually recognized that the interaction of the two approaches
into a cyclical model usually produce the most sound and comprehensive research. There
also exist other ways of conducting research which are not introduced here.
Again, whether deductive or inductive, both approaches must have the two above-
mentioned characteristics: (1) make sense and follow some logical thought; (2) use the
observation and data made during the research process.
Measurement and Research methods
It was said earlier that social research is about studying in a scientific way how people act
and interact in a society. Applied social research, it was also mentioned, is about
understanding a problem or an issue that affects communities.
Whatever the type of social research (exploratory, descriptive, explanatory) and the
scientific approach chosen to conduct research (deductive or inductive), research requires
measuring and observing human activity and human behavior.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
20/359
14
()
()
()
()
()
( )
(
)
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
21/359
15
Measurement and observation about human activity, behavior, mindset, and social
problems involve identifying the extent, frequency, or degree of this activity,
behavior, mindset, or problem. Social researchers must constantly ask themselves the
following: how frequently (not frequently, rather frequently, very frequently), how
seriously (not seriously, rather seriously, very seriously), how widespread (not
widespread, rather widespread, very widespread), .activities, behaviors, mindsets,
or problems are.
To answer these questions, qualitative or quantitative measurements are conducted.
Researchers decide which ones, quantitative or qualitative, are most appropriate
based on the research topic and the human resource and financial capacity available
for the research. Sometimes, it may be possible to conduct both qualitative and
quantitative measurements.
The selection of either quantitative or qualitative measurement for the research also
means the selection of different social research methods. The present volume
addresses this question. It introduces a variety of qualitative and quantitative research
methods that are used to measure human activities, behaviors, mindsets, or social
problems.
For conducting social research in a scientific way, researchers think not only about the
type of measurement (qualitative or quantitative) but also how to measure. The
success of a research project and its scientific value are primarily based on the choice
of the research methods and the quality of measurement.
The quality of measurement includes, but is not limited to, validity and reliability of
the measurement. These two technical considerations of measurement are complex
yet crucial for understanding how social research is conducted. They are introduced
here but will receive further development in the chapters.
To ensure validity of the measurement, it is important to ask oneself the following
question: does the way I measure this activity, behavior, mindset, or problem include
all the aspects of this activity, behavior, mindset, or problem that are measured for
this particular research project? There are other relevant questions to ask oneself
about validity of measurement but they are not introduced in the present volume.
To ensure reliability of the measurement, it is important to ask oneself the following
question: does the way I measure this activity, behavior, mindset, or problem produce
the same accurate and consistent results each time I measure this activity, behavior,
mindset, or problem? Similarly, there are other relevant questions to ask oneself
about reliability of measurement but they are not introduced in the present volume.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
22/359
16
() () ()
. . . ( )() (
)
(
)
(
)
[]
() ()
() () ()
()
() ()()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
23/359
17
This introduction has shown that social research is a scientific way of understanding
how people act and interact in society and how problems often emerge because of
peoples actions and interactions. Depending on the existing knowledge about a
specific topic, researchers may choose to study it in an exploratory, descriptive, or
explanatory way. This last way is the most challenging because it involves identifying
a link between two factors or situations and defining the nature of this link. If there
exist a causal link between two factors or situations, caution was given to avoid
establishing a causal link that determines a particular outcome.
Conducting social research as a scientific activity includes two necessary features.
First, the researcher develops broad and logical reasoning. Second, the researcher
interprets all observations and data from the research and frames them within thisbroader reasoning. Depending on how these two points are articulated by the
researcher, the scientific process can go from the general to the particular: this is the
deductive way. Or it can go from the particular to the general: this is the inductive
way. Lastly, as a scientific process, social research involves measurement of the
extent, frequency, or degree of an activity or problem. Social measurement is either
quantitative or qualitative. The decision by the researcher to opt for either a
quantitative or qualitative measurement depends on many factors such as research
topic, time, funding, human resources, . The scientific value of this qualitative or
quantitative measurement is based on quality criteria such a precision, validity, and
reliability.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
24/359
18
()
()
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
25/359
19
Glossary
Data: information collected during the research. Data cant appear as such
in a research report. Data is information about individuals, groups of
persons, or organizations while evidence is based on a set of data about
these individuals, groups of persons, or organizations. For example, Ko Myo
who lives in Shwe pyi thar Township lives on 800 kyats/day: this is data.
90% of all the persons surveyed in Shwe pyi thar Township lives on an
average of 800 kyats/day: this is evidence.
Data source: data produced through research can come from different
sources: observation, notes, interview tapes and transcripts, newspapers,
surveys and questionnaires, statistics, etc.
Data analysis: Data have to be analyzed, that means that the researcher (1)
enters all data into a computer, (2) seeks for evidence, and (3) examines
whether these evidence form general social patterns related to the
problem under study.
Evidence: as explained above, evidence is an aggregate of data about
individuals, households, or groups of individuals.
Fact: an observation that is objective and can be verified by researchers.
For example, contaminated water considerably increases the risk of
disease is a fact.
Finding (or research finding): the principal result(s) of a research project
presented in a report. For example, the research shows that villagersknowledge of the presence of the contaminated water in the underground
table did not prevent them from using it for most of the households
activities is a finding.
Attribute: a characteristic of a person or a thing. For example, male/female
is an attribute.
Category: a category is almost similar as an attribute except that category is
defined by the researcher for measuring a variable. For example, the
researcher defines the different levels of income that are relevant to the
purpose of a particular study: categories defined for levels of income willvary if the study is undertaken in the countryside or in Yangon.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
26/359
20
() () () ()
%
() () ()
() ()
( ) ()
() /
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
27/359
21
Variable: usually the problem or issue that is studied for research purposes. The
variable is to be measured quantitatively or quantitatively, or both. A variable can be
independent or dependent depending on the context of the study. For example, when
studying primary education, surface area of classrooms is an independent variable
that influences the ratio teachers/schoolchildren, here a dependent variable. BUTratio teachers/schoolchildren becomes an independent variable when researchers
examine how high or low enforcement of child-centered approaches, here a
dependent variable, is at primary school level.
Unit of measurement: the unit is defined by the researcher; it is the unit from which
information will be collected in order to measure the variable. In most cases, units of
measurement are individuals, households, or groups of persons legally or informally
established (see example of units of measurement in Chapter 2).
Hypothesis: it is a statement made by the researcher at the beginning of the project
regarding what he/she expects to observe about the link between two variables. The
expectation the researcher has about these observations is based on some theoretical
background.
Research question: it is also a statement about what the researcher expects to
observe during the research process; as opposed to a hypothesis, however, a research
question is not linked to a theory.
Sampling frame: a list or a set of assembled list of units composing a population from
which a sample is selected. For example, if a sample of students for a study of
graduates in Yangon division is selected from a compilation list of university graduates
in the division, all these lists put together are the sampling frame.
Sampling: a research method that researchers use to select a few people for study
and discover information about them that also apply to a larger of population that
cant be studied. Sampling is divided into probability and non-probability sampling
(both developed in this volume in separate chapters).
Indicator: measurement made about one aspect of a variable. Measuring an indicator
helps to understand how a variable changes throughout time. For example, a
researcher examines how deep a population in Mandalay division adheres toBuddhism (the variable being religious adherence). One of the possible indicators is
frequency at whichindividuals go to the pagoda.
Informant: an individual who has significant knowledge of the problem under study
and who is willing to tell what he/she knows about it.
Respondent: an individual who provides information by responding to a survey questionnaire
(different from informant, see above).
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
28/359
22
() () /
/
() () ( )
()
( )
( )
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
29/359
23
Stakeholder: individuals, or groups of individuals whose life, livelihood, or
professional activity is directly linked to the problem under study.
Quantitative research method: technique of collecting information based on large
number of informants. These techniques include survey and sampling.
Qualitative research method: technique of collecting information based on an
extensive interaction between the researcher and the informants. They include
Focus Group Discussion (FGD), In-Depth Interview, Participatory Learning
Appraisal, Participatory Action Research.
Livelihood: refers to the means of securing the basic necessities - food, water,
shelter and clothing- of life. It is defined as a set of activities, involving securing
water, food, fodder, medicine, shelter, clothing and the capacity to acquire these
necessities. Decent livelihood is achieved by working either individually or as a
group without having ones dignity being harmed by something or someone.
Well-being: refers to a situation characterized by health, happiness, and prosperity
for an individual or a community.
Social capital: refers to the networks, links, shared values in society that enable
individuals and groups to trust each other and so work together. The more social
capital one individual has, the more likely he/she is able to move socially upward.
Triangulation: refers to the use of two (or more) methods in a study in order to
check the results. Because much social research is founded on the use of a single
research method, and as such may suffer from limitations associated with that
method, triangulation offers the prospect of increased confidence.
Baseline study: refers to a descriptive survey covering a wide range of aspects
related to livelihood. The main purpose is to measure these aspects of livelihoods
for a population at a certain time. The data collected will serve as reference
materials for future surveys.
Problem tree: refers to a graphic tool mapping out a logical sequence of causes
and effects around a problem that is studied by researchers. The problem is broken
down into causes and consequences that may be studied as variables.
Reliability: is about whether the measurement of a variable is reliable or not. In
the context of a survey measuring religious practices, did you go the pagoda
during the last full moon? is a question for which answers will be more reliable
that the question about how many times have you been to the pagoda in your
life?
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
30/359
()
(Focus Group Discussion -FGD) (In-Depth Interview) (Participatory LearningAppraisal-PLA) (Participatory Action Research-PAR)
()
()
()
24
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
31/359
25
Validity: is about whether the measurement of a variable under study offers valid
information covering all aspects of this variable. In the context of a survey focusing on the
livelihood of a community, asking individuals their views about current politics would not
provide valid information for the focus of the research.
Bias: refers to a flaw in the research process that may lead to a wrong measurement of
the variable. There are various types of bias, which are described in the chapters of the
present introduction.
Anonymity: refers to a procedure that make it impossible for the researcher and the
reader of the research report to identify the findings with a particular individual or group
of individuals (different from confidentiality, see below).
Confidentiality: refers to a procedure that makes sure that only the researcher knows the
identity of the individual from whom information is obtained.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
32/359
26
() ( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
33/359
27
BASIC 1.DEFINING A RESEARCH PROBLEM
THE CHAPTER IS ABOUT:
The research steps to define a problem;
The problem trees as a tool for gaining better understanding of an issue, itscauses, and its consequences;
The definition of an issue or problem affecting communities in five questions:What, Who, Why, Where, and How?
______________________________________________________________________
This chapter introduces readers to the preliminary steps that researchers should follow
when defining a research problem. These are the first steps of the research design, which
will receive a comprehensive treatment in the intermediate level section of the manual.
RESEARCH AND EXPANDING KNOWLEDGE ABOUT AN ISSUE
When implementing social projects centered on an issue or problem, research producesnew knowledge about this issue or problem that impacts certain groups of the population.
This new knowledge often leads researchers to put into question what they think they
know already about this issue or problem.
Therefore, conducting social research is about two points. The first point is about
investigating the issue or problem in an open-minded way. The second point is about
gaining a deep understanding of the issue or problem.
When they investigate the issue or problem, researchers produce new informationabout
this issue or problem based on the data collected during the research process. This
information is new: this is not some information that can be entirely taken from the work
done by other researchers.
The research process helps the researcher acquire an informedview of the issue that is
under study. When the researcher disseminates the findings, more people can gain an
informedview of the issue or problem.
Very often, people are confused between the root causes and the consequences of the
problem. This happens because they lack an informed view of problem. By doing
research, the researcher can know more precisely what the root causes and
consequences of a problem are.
Researchers study Child Drop-out in Hlaing Township. They mayinclude some statistical data provided by the Ministry ofEducation about drop-out rate nationwide or specifically inHlaing Township.But this data is not sufficient. It is crucial that theresearchers also include data collected from the field workconducted in Hlaing Township. This data is the new informationthat organizations tackling child drop-out may use for their own
projects.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
34/359
28
-
( Problem tree )
( ) _____________________________________________________________________________________
[
]
() () () () ( )
( )
()
" ( ) "
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
35/359
29
DRAWING APROBLEM TREEEven before drawing a problem tree, it is recommended to read available
reference materials about the problem investigated. The more materials that are
read, the better. This may help one have a clearer view of the problem.
One effective way to try to identify the root causes and consequences of aproblem is to draw a problem tree:
(1) when drafting the project proposal or at the beginning
of the project;
(2) at the end of the project.
The first problem tree presents the level of understanding of the specific issue.
Because people at the beginning of the project may not have a thorough
knowledge of the issue, its root causes, and consequences, the problem tree is not
very precise. It may also include preconceived ideas.
After the research project is complete, a second problem tree has to be drawn
again. At this time, it is expected that the new problem tree will have much more
information than the first problem tree: For example, the number of causes and
effects about the issue will be larger. Also, all preconceived ideas hopefully will
have disappeared!
First problem tree Second problem tree
The second problem tree drawn after research usually offers a good summary of the
research findings. In some cases, it may be a good option to place this second problem
tree in the final research report.
The root causes of the issue andthe impact of the issue on thecommunities are not wellidentified. The knowledge of the
researchers is limited. Some ofthe root causes and impact canoften be blurred and mistakenfor one another. The problemtree before research remainsquite vague and featuresgeneralities.
The root causes and the impact of the issueare clearly identified and rightly placedin relation to the issue that is studied.Furthermore, the number of root causes and
the consequences linked to the issue orproblem under study is much larger thanthose in the first problem tree. Theresearch process may also question some
preconceived ideas that were featured inthe problem tree before research.
After research
Consequences
Issue or Problem
Root Causes
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
36/359
30
Problem Tree
Problem Tree ()()()
Problem Tree Problem Tree
-
.
( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
37/359
31
Below is an example of a problem tree created after research is conducted. The
problem is related to health issues and does not directly address social issues. But it is
nevertheless instructive because it shows more easily the link between the problem, its
causes, and its effects or consequences.
It is also important to note that after research, the problem tree may also identify some
primary effects or consequences, and some secondary (indirect) effects or
consequences.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
38/359
32
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
39/359
33
WHAT,WHO,WHY,WHERE,AND HOW?
Drawing the first problem tree is also a preliminary step in answering five
questions. Answering these five questions is essential in implementing a project and
defining the research component of the project. More generally, these five questions
are the basic questions that need to be answered in the preliminary phase of theresearch design. A comprehensive treatment of Research design is given in the
intermediate level section of this manual.
What?
Answering the What? question means to define the issue or problem addressed.
One big difficulty for local CSOs and NGOs is the definition of an issue or a
problem that can be realistically addressed based on the human resources and
financial capacity of the organization.
Issues like poverty, livelihood, or discrimination are too big if they are
addressed by only one NGO or CSO. All these issues have a wide range of causes
that cannot be resolved by one specific project implemented by one single
organization.
This, however, should not prevent researchers or organizations from gaining a broad
understanding of the problem, its root causes, and consequences.
Even though the organization may address only one root cause of the problem as its
project, there are benefits in doing research on its many root causes.
Who?Another crucial issue is defining the type of population that the research team
examines. This issue is addressed several times in this manual because it depends on
whether the research team uses quantitative data based on surveys or qualitative databased on qualitative interviews.
The Who? question also means that the research team has to decide whether the issue
or problem is linked to (1) individuals, (2) households, or (3) groups of people(legal
associations, self-help groups, etc). It would be wrong for a research team to
investigate an issue by using two or three of the three categories mentioned above at
the same time and for the same project. Only one category should be selected for a
specific topic.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
40/359
34
?
?
"" ( ) () ( )
( )
()
?
( ) ( ) ( ) ( ) ( )( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
41/359
35
After deciding whether the research issue is linked to individuals, households,
groups of people, etc., the research team defines more precisely the criteria
associated with the category of population which is studied. For instance, if the
research project studies individuals, the team may select the following criteria: age
of the individuals, gender of the individuals, educational background of individuals,
etc The decision to include one or more of these criteria depends on how each
criterion is relevant to the research topic.
The same is valid whether the research project studies households or groups of
people, although the criteria may be different.
Why?The research team has to answer the Why? question. This means: why does the
research team focus on this issue or problem? The main reason, of course, is that
this issue affects the communities that are targeted by the project. The research
team may have some ideas about why the issue or problem emerges at a specific
time and in particular place or is already so acute at a specific time and in particular
place. In this case, the research team conducts research in order to check whether
these ideas are valid and how solutions to the issue can be found.
A research team examines the issue of child drop out. Thepopulation selected for the research is a large number ofchildren who dropped out of school. These children are all (1)individuals. The team may contact parents, teachers, and schooldirectors for the study. But the population that is exclusivelyexamined here consists of drop-out kids.
For a research project focusing on access to water, it isrelevant to study the (2) household. Because access to water isshared within one household or one family, it makes sense tostudy this issue among households only.
A research project studies how community forestry is managedamong (3) self-organized groups of villagers in a mountainousarea. The research team contacts heads of villages, villagers,government authorities, and other stakeholders. The research,however, uses groups of villagers as category of population that
is under study because it studies how these different groupsinteract with each other in order to manage community forestrycollectively and sustainably.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
42/359
36
( )
( ) ( )
( )
-
?
- ( ) ()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
43/359
37
When answering the Why? question, concrete examples of howthe issue affects these
communities should be given. There may also be many other reasons that explain why
an organization conducts projects in a certain field of activities and why the
organization conducts research about this problem before implementing the project
itself. These reasons have to be clearly stated in any documents produced by the
organization.
Where?Myanmar is a very diverse country with many different geographical, social and
economic contexts. The team in charge of conducting research on a specific issue has
to decide where the research project is going to be conducted. Depending on the
organizations background and scope of activities,the location of the projects will be
different. The selection of the place also has to consider the language issue: For
example, in areas where the Myanmar language is not widely spoken, can the research
be conducted using Myanmar-language materials and Myanmar-language interviewers
only? Translating survey materials from the Myanmar into the local language and
hiring local translators will probably be two issues to seriously consider when the
location of the project is decided.
How?
Answering the How? question is primarily about whether the research will rely on
quantitative data, qualitative data, or both. All this will eventually depend on the
human resources and the financial capacity available to the organization. In terms of
human resources, it depends on the research skills and experience of those conducting
the research.
After deciding whether to use quantitative research, qualitative research, or both, the
research team has to decide how to collect the data (i.e. which research methods will be
used).
TO SUM UP,THE MAIN POINTS TO REMEMBER ABOUT THIS CHAPTER ARE:
How conducting research at the beginning of a project must expand knowledgeabout an issue or problem that affects communities;
The importance of drawing a problem tree at the beginning of a project and atthe end;
The importance of answering the five questions: What, Who, Why, Where, andHow?
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
44/359
38
?
?
( )
?
() ()
() ()
( )
()
( Problem Tree )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
45/359
39
BASIC 2.QUANTITATIVE AND QUALITATIVE RESEARCH
THE CHAPTER IS ABOUT: The main features of quantitative research;
The main features of qualitative research;
The difference between quantitative research and qualitative research;
The notion of a variable linked to a problem or issue;
The use and measurement of a variable in quantitative and qualitative research.
_____________________________________________________________________
There are three broad types of investigation for conducting social research. Selecting
the suitable type depends on the level of knowledge about the issue or problem
addressed. Sometimes, the knowledge of the issue and problem under study is very
limited. Sometimes, it is quite advanced. Social research can therefore be conducted
through the following three approaches:
-Exploratory, meaning that the researcher tries to gain a preliminaryunderstanding of the situation under study, because there is no or little
preexisting knowledge about the problem under study;
-Descriptive, meaning that the researcher tries to measure the issue under study
so as expand the existing knowledge about the problem under study;
-Explanatory, meaning that the researcher tries to find out some connections
between different aspects of the situation under study.
In many cases, research projects can combine two or three of these different aspects.
All this depends on the nature of the research topic, the level of knowledge about the
topic, and the human and financial resources available to the project. Also, it should
be remembered there are no clear-cut boundaries between these three aspects: in
some studies, these aspects can overlap.
Lets take the example of a research on Karen communities living
in the Ayeyawaddi Division.
An exploratoryaspect of the research is about identifying Karencommunities in the overall population of the division.
A descriptive aspect of the research is about the level of landand food security typical in these Karen communities.
An explanatoryaspect of the research is about finding out some ofthe causes linked to poverty among the Karen communities.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
46/359
40
.
()
_____________________________________________________________________________________________
-(Exploratory)
() -(Descriptive) -(Explanatory) ( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
47/359
41
For each of these ways of conducting social research, data may be collected by using
qualitative research methods. This means using interviews with the people directly or
indirectly linked with the issue under study.
The data may also be collected by using quantitative research methods. This means
surveying the people directly linked with the issue under study.
Research can therefore be qualitative and quantitative.
QUANTITATIVE RESEARCH
The examples given previously show that quantitative research produces statistical
data. Quantitative research focuses on numerical measurement. It frequently uses the
survey as its main research method.
Final research reports usually contain statistical data and percentages that more easily
catch the attention of the reader.
If researchers study livelihoods of Karen communities in theAyeyawaddi Division (like in the previous example), they canproduce both qualitative and quantitative research.
Quantitative research with an exploratory purpose produces a
percentage of the Karen population among the entire population ofthe Ayeyawaddi division and assesses its poverty level.Qualitative research with the same exploratory purpose givesinsights into whether livelihood difficulties among Karencommunities are different from those of other communities in theAyeyawaddi division or not.
Quantitative research with a descriptive purpose producespercentages of levels of land tenure and food production amongKaren communities.Qualitative research with the same descriptive purpose givesinsights into the range of difficulties encountered by Karen
communities for attaining land and food security.
Finally, quantitative research with an explanatory purposeattempts to determine statistically if a link between land andfood security does exist among these Karen communities.Qualitative research with the same purpose investigates whetherthese Karen communities perceive a possible link between land andfood security in the context of their villages.
There are therefore many ways in which a research project can becarried out: there is no single way to conduct research on aspecific topic.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
48/359
42
()
(
QUANTITATIVE RESEARCH)
(statistical data)
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
49/359
43
QUALITATIVE RESEARCH
In contrast to quantitative research, qualitative research does not focus on numerical
measurement. The examples given previously show that qualitative research tries to
illustrate the deep meaning of the problems under study. Qualitative research typically
examines individual cases. It frequently uses interviews and field observations withoutstrict numerical measurement.
Qualitative research eventually produces findings that are richer in meaning and
complex in details. Final research reports usually contain lots of observations,
testimonies, and also interpretations of these observations by the researcher.
VARIABLE
Quantitative research and qualitative research involve collecting information related to
a problem or an issue under study. Very often, however, final research reports are
vague and do not contain any new information about the problem or issue that wasexamined. One of the main reasons for this lack of information is that the researcher
did not think about howthe problem or issue could be broken down into one or two
variables. Also the researcher did not think how such variables could be measured.
At the beginning of a research project, it is important for the researcher to consider how
the problem under examination can be broken down into variables that will be
measured.
A variable is anything that can have different values. Height, weight, age, race, and
attitude are variables because there are different heights, weights, ages, races, and
attitudes. By contrast, if something cannot vary, then it is referred to as a constant.
Broadly speaking, quantitative researchers usually work with a few variables and many
cases, whereas qualitative researchers usually rely on a few cases and many variables.
In the example given previously about the livelihood of Karencommunities in the Ayeyawaddi Division, the problem of livelihoodis very broad. It is difficult to study as such. The researcherhas to decide which aspects of livelihood will be studied. Theresearcher makes this decision after drawing the problem tree(discussed in a previous lesson) and identifying the root causesand consequences of this problem.
It was suggested previously that land security and food securityare important aspects of livelihood. These two aspects (landsecurity and food security) can be chosen by the researcher asvariables under examination.
Another variable for the research topic of livelihood could beincome generation.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
50/359
44
(
QUALITATIVE RESEARCH)
(
VARIABLE)
()
(variable) (constant)
() (problem tree)
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
51/359
45
Besides values, a variable has alsoattributes. Attributes are characteristics that describe
a person, an object, and an abstract thing.
For onevariable under study, there are manyattributes. If is often important for
researchers to use different research methods to study a variable. If financial and
human resources allow and time is sufficient, researchers would benefit from
investigating a variable in quantitative and qualitative ways.
Here is a list of variable with some suggested attributes. It is the researchersresponsibility to define these attributes.
VARIABLE ATTRIBUTE
age young, middle aged, old
gender female, male
occupation plumber, journalist, driver
nationalityMyanmar, Japanese,
Australian
social class upper, middle, lower-class
political viewsconservative, liberal, non-
partisan
interest in politics low, moderate, high
For the research on the livelihood of Karen communities in theAyeyawaddi Division, the two variables suggested (land security
and food security) can have different values: land security and
food security may be low, moderate, or high.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
52/359
46
(attributes)
( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
53/359
47
Very often a research project will investigate more than one variable, usually two
variables and their possible link. Then, the purpose is explanatory. This means that the
research project tries to clarify how one variable influences another variable.
Examining the link between two variables leads to studying them in the following way:
Independent variable: a variable that influencesthe other variable
Dependent variable: a variable that is influencedby the other variable
If a research project investigates the link between two variables, researchers have to
identify which variable is independent, and which variable is dependent.
Lets take the case of a study trying to find a link between education level and fertility.
The study reported that because fertility depends on the education level (and not the
other round) fertility is the dependent variable and education level is the independent
variable.
The research findings are as shown below.
The population with a low education level was found to have high fertility. In contrast
to this, the population that had a high education level was found to have low fertility.
Although having many different aspects, two main attributes aboutland security are: land tenure and landlessness. Land tenure meansthe link between a plot of land and a person who has claim on it.Therefore land security may be poor, moderate, or high dependingon how strong the link between the land and the person is. By
contrast, landlessness relates to a person who cannot make anyclaim on any land.
Thus a research project that takes land security as the mainvariable has to examine the two main attributes of this variablewhich are land tenure and landlessness. The researcher may alsoexamine other aspects of land security.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
54/359
48
(Explanatory)
(Independent variable)
(dependent variable)
( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
55/359
49
Very important note: these are the results for a specific context. In a context and
society different from this specific study, the influence of education on fertility might
be different.
CATEGORIES
It was seen in the previous section that variables can be either independent or
dependent. This distinction between independent and dependent, however, does
not help the researcher measure variables.
Measuring a variable means creating categories for classifying persons, objects, or
abstract things about which information is collected. In many cases, categories are the
same as attributes that define the variable (as featured above in the table about age,
gender, etc.). In some cases, however, categories have to be created in order to make
possible the measurement of the variable and, most notably, its numerical
measurement.
To define age as a variable only with attributes like young, middle-aged, or old
is too limited for a research project. The researcher has to create categories to define
what young, middle-aged, and old mean in the context of the research. Young
may mean between 10 and 21 years in one research project or between 10 and 25 years
in another research project. The researcher must decide which categories better suit the
objectives of the research.
In short, the researcher has to remember that attributes are just the general features of a
variable. Categories, however, are created for the measurement of the variable.
There are tworequirements for the creation of these categories:
These categoriesshould be exhaustive. The number of categories has to be aslarge as possible because all the persons, objects, or abstract things about which
information is collected must be classified within these categories.
These categoriesshould be mutually exclusive: every person, object, or
abstract thing about which information is collected must fall into only one
category. This means that the categories that measure a variable can notoverlap.
When the researcher creates categories for measuring a variable, these categories can
be names, numbers, or intervals. These different types of categories mean that the
variable has different names.
When categories of the variable are only names, the variable is called a nominal level
variable.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
56/359
50
(
CATEGORIES)
() () (categories) ( )
()()
(exhaustive) ()()
(mutually exclusive) () ()
() ()
(nominallevel variable)
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
57/359
51
When categories of the variable are names, labels, and can be ordered in some kind of
ranking, the variable is called an ordinal level variable. The different categories also
have names (like the nominal level variable just mentioned), but ordinal level variables
can be ordered as having lower or higher values in relation to one another.
When categories of the variable under study have names, can be rank-ordered, and
represent standard intervals of measurement, the variable is called an interval level
variableor a ratio level variable. The only difference between the two terms interval
level and ratio level is that ratio level has an absolute zero.
A research project examines the issue of migration into Yangon andtries to determine where the migrants come from. Place of originis one of the variables under study; the researcher has to createdifferent categories that refer to geographical places. Thecategories for the variable place of origin are geographical
names: Rakhine State, Ayeyawaddi Division, Bago Division,etc
This way, the researcher can determine a percentage of how manypeople come to Yangon from Rakhine, how many come from Ayeyawaddior Bago. This is the kind of data that quantitative researchproduces.
For qualitative research, the researcher can interview Rakhinemigrants and examine their reasons for migrating to Yangon. Theresearcher can then compare these reasons with those mentioned bymigrants coming from the Ayeyawaddi and Bago Divisions.
This is an example of a nominal level variable.
A research project examines youths support of political parties.The categories created for assessing political support as themain variable of the study are the ones suggested here: (1) nosupport; (2) moderate political support; (3) active support; (4)formal membership; and (5) active membership.
As opposed to the previous example in which place of origin isonly a nominal level variable, here, in this example, thevariable has different values attached to each of the categoriesgoing from no support to strong political support (activemembership).
This is an example of an ordinal level variable.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
58/359
52
(ordinal level variable)()
(interval level variable) (ratio level variable)
- () ()() () ()
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
59/359
53
UNIT OF MEASUREMENT
A unit of measurement is a unit about which information is collected for a research
project. A unit of measurement will help to measure the variables that research projects
study. Units of measurement are the people, objects, or abstract things that researchers
will observe and describe during the research project. Typically, the unit ofmeasurement in social research is the individual person, but it may be a social group,
an organization, etcThe unit of measurement chosen by the team answers the Who?
question explained in the previous chapter: it is about choosing the type of population
most relevant for the variable that is measured by the research project.
Household income is often taken by research projects as anindependent variable because it influences other variables. It hasthe following attributes: low, middle, and high. But theresearcher has to define what low, middle, and high means in the
context of the research project. The researcher has to createcategories which in turn help measure income.
For instance, a first category is for incomes strictly below 1lakh; a second category is for incomes between at least 1 lakh andstrictly below 5 lakhs; a third category is for incomes between atleast 5 lakhs and strictly below 10 lakhs, etc It is important to
note that the categories are created for only a specific project.
The number of categories and what each category means is veryoften different from one research project to the other, unless theresearcher intends to compare two different contexts with the same
categories.
This is an example of an interval level variable.
If a research project focuses on the role of civil societyorganizations in Myanmar, the unit of measurement is the civilsociety organization.
If a research project examines the livelihood of Karen communitiesin the Ayeyawaddi Division, the researcher might find it relevantto select the Karen household or a Karen family living in onehouse as a unit of measurement.
If a research project investigates employment opportunities foryoung Myanmar graduates, the unit of measurement is the individual(in this case, young Myanmar graduates).
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
60/359
54
(
UNIT OF MEASUREMENT)
() ()
()
( )
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
61/359
55
Choosing the right unit of measurement is crucial for conducting good research and for
the validity of the research.
RELIABILITY AND VALIDITY OF RESEARCH
The quality of a research project includes, but is not limited to, the validity andreliability of the findings produced. High validity and reliability is achieved when units
of measurement, categories, variables, and research methods are adequately selected by
the research team.
To increase the validity of the findings, it is important for the researcher to ask
him/herself the following questions: Are the variable(s), categories, and units of
measurement adequately defined and selected with regards to the problem that is to
measured or investigated? In other words, can the researcher expect that his/her
definition of the variable(s) as well as his/her selection of categories and units of
measurement will produce data that adequately addresses the problem investigated?
To increase reliability of the findings, it is important for the researcher to ask
him/herself the following question: Does the definition of the variable(s) and selection
of categories and units of measurement enable the researcher to obtain consistent data
throughout the duration of the research?
Issues of validity and reliability will recur for the selection of the research method(s)
and the adequacy of the selected research method(s) with the topic investigated.
TO SUM UP,THE MAIN POINTS TO REMEMBER ABOUT THIS CHAPTER ARE:
Quantitative research requires the identification of variable/variables that aregiven some numerical measurement by the research;
Qualitative research may also measure a variable, but not in a numerical way; italso investigates problems and concepts, evaluates their meaning, depth, and
physical manifestations;
Both quantitative and qualitative research require the precise definition of whatthis/these variable(s) mean(s) in the context of the research project;
Both quantitative and qualitative research usually require the identification ofindependent variable(s) and dependent variable(s);
Depending on the variable(s), the categories created for measuring thevariable(s) will be: (1) names, (2) ranking/degree linked to the variable, or (3)
regular intervals on a measurement scale;
Measurement of variables is made on the basis of units of measurement, themost commonly used being (1) individuals, (2) households, or (3) social groups.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
62/359
56
( )
( )
() ( )
-
/
( )
( ) ()()/()
() () ()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
63/359
57
BASIC 3.BASELINE STUDY
THE CHAPTER IS ABOUT:
The importance of the baseline study as a means of collecting preliminary
information before implementing a social project; The collection of baseline information centered around five types of capital
(Human, Natural, Social, Physical, and Financial); The selection of baseline indicators which can be used for monitoring the
progress of a project.
______________________________________________________________________
BASELINE STUDY:GENERAL CONTEXT AND DEFINITION
A crucial problem in Myanmar (and, very often, elsewhere) is the lack of reliable
information. When beginning a social project, CSOs and NGOs often have very
little reliable information for implementing their projects. This can be even more
difficult when a project is implemented in the States of Myanmar where thepopulation is not predominantly Bamar. In these different States, the population has
a different culture, society, and history than in the Division of Myanmar. If the
project managers have a prior knowledge and understanding of how things work
there, the project is more likely to succeed in reaching its objectives.
It is particularly important that NGOs know the social and economic contexts in
which the projects are implemented. Use of local resources, farming techniques,
social structure of local communities, and weather conditions are crucial factors that
should be known before any project is implemented. These contexts or factors
should inform the way the project is designed and the way the project tries to reachits objectives. If this information does not exist before starting the project, it should
be collected as part of a Baseline Study when the project starts.
The Baseline Study is not a research method: it uses different research methods to
address a particular problem or issue.
The purpose of the Baseline Study is to give detailed and reliable information
that can be used for monitoring the progress of a project during its
implementation and assessing its success when it is completed. Comparing datacollected at the beginning of the project and at the end makes it possible to
measure how much change has happened during the implementation of the
project.
When measuring how much change has happened during the project, the researcher
should try to determine the extent to which the change occurred because of the
project was implemented and the extent to which other factors may have influenced
also the change.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
64/359
58
.
()
_______________________________________________________________________________________
()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
65/359
59
The example below shows a graph made as part of a baseline study focusing on
livelihood. The project implemented by the organization seeks to positively change the
livelihood of the involved communities. The results of this change will be seen in the
sustainable livelihood outcomes and the increased security of the household at various
levels (see box in the right part).
Adapted from: http://www.fao.org/docrep/003/x9371e/x9371e11.htm
To measure this positive change, however, the research team must first understand the
context (box on the left side). The team must also understand the five different types of
capitals which influence the life and future of the household (lower part of the
livelihood strategy in the center). The definition of capital, in this context, is all the
assets available to a household that the household can use to increase its welfare.
By conducting a Baseline Study, the research team can improve understanding of both
the context and the different types of capital as well as all other aspects of the
livelihood strategy shown in the above graph.
HOW TO CONDUCT A BASELINE STUDY
A Baseline Study should be conducted as illustrated below:
Prepare a Baseline Plan
Conduct the Baseline Study according to the Baseline Plan
Analyze the collected data
Write a Baseline Report
HOUSEHOLD
Consumption(eating, buying
goods, etc.)
Production &
Income
Human Capital Natural Capital Social Capital Physical Capital Financial Capital
Social Activities
(helping the community, self-help
groups, associations, etc.)
Livelihood Strategy
Natural Resources
Institutions
Infrastructure
History
Social Context
Economic Context
Demography
Sustainable
Livelihood
Outcomes
Context
Security of:
Food
Health
Water
Shelter
Education
Land
Community
cohesion
Personal safety
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
66/359
60
( )
http://www.fao.org/docrep/003/x9371e11.htm
() ( )
http://www.fao.org/docrep/003/x9371e11.htmhttp://www.fao.org/docrep/003/x9371e11.htmhttp://www.fao.org/docrep/003/x9371e11.htm -
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
67/359
61
The Baseline Plan is drafted by the researcher who clarifies what information is needed
for the Baseline Study and what indicators are used to measure change. But before
explaining what the Baseline Plan consists of, lets examine the type of information
needed.
Baseline
General
information
Themes for each type
of capital
Possible indicators for each type of capital
Human
Capital
Labour(1) Rate of hired labour; (2) Rate of small landholder; (3) Emigration ofpopulation; (4) etc.
Health and nutrition status(1) Main food items for daily diet (2) Ratio height/weight; (3)
Immunization against disease; (4) Ratio population/doctor; (5) etc.
Skills and knowledge(1)Male/Female literacy; (2) Apprenticeship (3) Attendance to training;(4) etc.
Natural
Capital
Access to land (1) Irrigated land; (2) Non-irrigated land; (3) Forest land; (4) etc.
Water(1) Number of wells within the village; (2) Free access to wells; (3)Annual rainfall; (4) Irrigated area for farming; (5) etc.
Renewable and non-renewable resources
(1)Soil capability; (2) Rivers for fishing, washing, and other activities;(3) etc.
Flora, forest, and wildlife(1)Free access to forests in proximity of residence; (2) Negotiated use of
forest resources; (3) Sustainability of forest management: (4) etc.
Social
Capital
Social network(1)Mills operated collectively among villagers; (2) Engagement with local
authorities; (3) etc.
Social trust (circle of trusted
people)(1) Self-help groups; (2) Religious associations; (3) Non-religious
associations; (4) etc.
Physical
Capital
Housing(1) Thatched house; (2) Wooden house; (3) Brick house; (4) Plastic tent;
(5) etc.
Vehicle & Infrastructure(1) Distance from home to work place; (2) Ownership of a vehicle; (3)
Condition of road infrastructure; (4) etc.
Equipment (1) Engine; (2) Farm mechanization; (3) etc.
Livestock (1)Cattle; (2) Poultry; (3) Pigs; (4) etc.
Financial
Capital
Savings (1) Cash in household; (2) Debt of household
Access to funding (1)Access to loans; (2) Access to micro-credit; (3) etc.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
68/359
62
()()()()
()()/ () ()/ ()
()/() ()
()
() () ()()
()()() () ()
() () ()
()() ()()
() () ()
()
() () () ()
()() ()()()
() ()()()
() ()()
()()()()
()()
() () ()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
69/359
63
IMPORTANT NOTE:
All types of indicators above in the table are mentioned for the purpose of
clarification only. They are not all necessary needed for a Baseline Study.The selection of indicators is crucial for the research process. An indicator is one
aspect of a variable (see Chapter on Qualitative and Quantitave Research for
definition of a variable). Because a research team studies how this variable varies,the team has to consider which aspects of the variable are to be measured. An aspect
of a variable that can be measured is an indicator.
Baseline
Plan
Indicators
Selected
Data Collection
Methods
Location of Data
Collection
Time needed for
Data Collection
Human Capital
Natural Capital
Social Capital
Physical Capital
Financial Capital
Literacy is a variable. A research team has to decide how tomeasure literacy. This means that the team has to select arange of indicators that the team thinks can measure theliteracy variable.If the research team measures level ofknowledge among communities in remote mountainous areas, theindicators selected might be: (1) Level of school completed;
(2) Reading proficiency; (3) Writing skills; and (4) Basic Mathproficiency.
Note that the indicators for measuring knowledge can bedifferent if the research team works in remote areas and inurban areas.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
70/359
64
(()
)
)
() () ()()
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
71/359
65
Selecting the right indicators for the Baseline Plan and, more generally, for the
Baseline Study is very important. These indicators can be used by the research
team to monitor the evolution of the socio-economic context during the project
implementation. They can also be used at the end of the project when an
assessment is made to determine whether the project has reached its objectives.
The next section gives further explanations of this issue.
A Baseline Study most often is based on quantitative data. But it can also rely on
qualitative data. Again the selection of one or other type of data depends on the
research topic and the human and financial resources available to the project. If
possible, the research team should try to use both quantitative and qualitative data
for the Baseline Study.
QUANTITATIVE ASPECTS OF THE BASELINE STUDY
Among the different indicators selected by the team, three indicators are taken
below as examples. These are examples only given for clarifying purposes. A
Baseline Study usually includes more indicators.
One indicator that can be chosen at the beginning of the project is the percentage of
farmers understanding the newly-revised land law. Understanding here could mean the
amount of right answers given by the farmers to a questionnaire about this new law.
The study (see graph below) shows that only 17% of the farmers of the three Divisions
that are part of the project understand the newly-revised land law at the beginning of
the project.
An organization implements a project on land security in threeDivisions of Myanmar. The purpose is to increase opportunitiesfor land security among small landholders. At the beginning ofthe project, the research team discusses which indicators ofland security will be used as baseline information. Theseindicators can then be used by the research team to monitor theprogress of the project.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
72/359
66
(Baseline Plan)
()%
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
73/359
67
(the time intervals shown here are only for clarification purposes;
any project has to decide which time intervals are the most relevant)
If the organizations human and financial resources allow, researchers could possibly
ask the same question to the same number of farmers at the projectsmid-term. Then
the researchers could assess whether there is any progress. At the end of the project,
this could be done a third and final time. By comparing the percentage at the beginning
and end of the project, the indicator could show whether some progress is made on this
aspect of the project.
A second indicator that can be chosen is the number of self-help groups that were
created by the small landholders during the course of the project.
The field survey collects information about the second indicator at the beginning of the
project. The Baseline Study shows (see graph below) that only two self-help groups
exist at the beginning of the project.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
74/359
68
(
)
()
T() () ()
( (-)
TT
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
75/359
69
The data collected at the end of the project for the same indicator will show whether
there is some progress made on this aspect of the project in the three Divisions of the
country selected for the implementation of the project.
A third indicator that could be chosen is the number of applications submitted for landregistration.
The field survey collects information about this third indicator at the beginning of the
project. The Baseline Study (see graph below) shows that 20 applications were
submitted for land registration.
The data collected at the end of the project for the same indicator will show whether
there is some progress made on this aspect of the project.
The research team analyses the data both at the beginning and at the end of the project.
The team has to consider how much of the progress is directly related to the activities
implemented by the project. It also has to consider other factors which might have
influenced change as shown by the indicators.
QUALITATIVE ASPECTS OF THE BASELINE STUDY
Qualitative aspects of the Baseline Study focus on information that is not easily
understood by simply conducting quantitative research. Social structure of families,
farming techniques, or mode of subsistence may be best understood by conducting
qualitative research.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
76/359
70
-
()
()
-
() ( )
T
T
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
77/359
71
In this case, literature review, focus group discussions, and key informants interviews
all are research methods that produce valuable baseline information.
TO SUM UP,THE MAIN POINTS TO REMEMBER ABOUT THE CHAPTER ARE:
A Baseline Study serves to collect preliminary information about communitiesthat benefit from a social project;
Some of the information collected during the study should be used as indicatorsthat help the researchers monitor how much change happens during the project
implementation;
By collecting data about these indicators at the beginning, middle, and end of
the project implementation, the researcher assesses the extent to which thechanges that occurred during the project implementation are directly related to
the activities conducted by the project.
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
78/359
72
-
7/23/2019 Inya.BilingualManualSocialResearch Methods.pdf
79/359
73
BASIC 4.SURVEYRESEARCH(1)
THE CHAPTER IS ABOUT:
The use of survey research for collecting data;